Abstract
The purpose of the present paper is to show an approach to the suboptimal control policy for controlling the nonlinear stochastic control system.
Specifically, the so-called continuous-discrete nonlinear stochastic control system is treated in this paper, because this type of control system seems to be important in practical situation. In this stochastic control system, the nonlinear system dynamics operate continuously and the non-linear observations are taken at the preassigned discrete time points.
First, applying the quasilinearization technique, the state estimation algorithm is proposed.
Then, using this estimation algorithm, a suboptimal control policy is proposed with the discrete type criterion function, instead of the criterion function of the continuous type, because of the discrete observations.
Then, a selection procedure is proposed for determining the unknown parameter included in this suboptimal control policy and, using this procedure, our suboptimal control policy is established.
Finally, the computer simulational experiments are shown, using our suboptimal control policy, in case of the several exemplary control systems.
Our policy and algorithm seem to be useful for the practical applications, so far as our experimental results are concerned.